精准农业土壤采样密度研究——以上海精准农业试验示范基地为例

通过土壤密集采样识别农田土壤物理生化性状的空间差异,构建土壤与农作物产量关系模型,用间接方法提高土壤性状信息的分辨率并定量评价这些信息的可靠性非常必要。在评价多种插值方法的基础上,对中国科学院上海精准农业示范基地每公顷3个样本、等距离系统抽样的280个土壤有机质样本采用克立格方法进行插值,估算样本之间土壤有机质含量和分析估算的可靠性,并与距离倒数函数插值法进行对比。对几个减少样本数量的方案进行分析,提出在满足精准水稻种植管理分区要求前提下可减少1/2原有土壤采样样本,并给出适于水稻精准种植土壤采样设计的逐步优化方法。

Study on soil sampling for precision agriculture-A case study from Shanghai Experimental Demonstration Base

It is significant to explore the spatial variation of soil biochemistry characteristic by densely soil sampling,to construct the mathematical models for revealing the relation between crop yield and its environment for precision agriculture,to raise the resolution of soil maps through indirect methods and evaluate its reliability.Based on the evaluation of interpolation methods,this paper adopts Kriging interpolator to produce soil map of organic matter in Shanghai Base,using 280 soil samples of systematic sampling of 3 samples per hectare.The comparison between Kriging and inverse distance function interpolating is carried out by using the same data set,showing that the Kriging interpolator is better than the latter in this situation.This study also analyzes the scenario of producing soil maps by reducing the amount of samples and gives the relevant evaluations.Based on these analyses,the authors present the procedure of optimal design of soil sampling for precision agriculture of rice growing.